Chapter 6/7 - Observational Methods, Measuring & Recruiting

Question 1

quantitative methods

Answer 1

• Unit of analysis: numerical data
• Method of analysis: statistical analysis
• ex. surveys, systematic observation

Question 2

pros and cons of quantitative methods

Answer 2

• Pros:
• excellent for testing hypotheses
• Allow for effects and differences to be measurable and tangible
• Allow us to determine if differences are likely to be due to chance – helps us draw valid conclusions
• Cons:
• Limits understanding of behaviours and thoughts to numerical representations of them
• Validity of turning thoughts and behaviours into numbers?

Question 3

qualitative methods

Answer 3

• Unit of analysis: personal experience and open-ended responses
• Method of analysis: interpretation of responses
• ex. focus groups, naturalistic observation

Question 4

pros and cons of qualitative methods

Answer 4

• Pros:
• Excellent for generating hypotheses
• Allow for in-depth verbal understanding of human behaviour
• Escapes issues related to distilling thoughts and behaviours into numbers
• Cons:
• Conclusions drawn may be due to random chance or personal biases
• Can be extremely time-consuming and difficult to interpret

Question 5

naturalistic observation

Answer 5

• Obtain/gather/record information from people in the field
• Can sidestep ethically problematic experimental manipulations
• can be difficult as researchers are unable to control the setting
• 2 types: participant observation and concealed observation

Question 6

participant observation

Answer 6

• one type of naturalistic observation
• The observer is participating along with the people being observed
• Purpose is made known to people in the group
• Problems: researcher may lose objectivity; reactivity -> the fact that they know they’re being observed influences the way they behave

Question 7

concealed observation

Answer 7

• one type of naturalistic observation
• Either the observation itself is concealed, or the purpose of the observations in concealed
• Ethical considerations may be problematic
• helps avoid reactivity

Question 8

systematic observation

Answer 8

• Observations of specific behaviours in a more controlled setting (ie. lab setting)
• Coding is often introduced to categorize what it being observed
• can be hard to achieve inter-rater reliability; reactivity may be an issue if researcher/video camera are visible

Question 9

case studies

Answer 9

• Very common in (but not exclusive to) clinical psychology
• Studying specific case (person, school, neighbourhood, etc.) in great detail - usually because the case is noteworthy or rare
• one type of case study: psychobiography (using psychological theory to explain the life of an important individual)
• problem: results cannot be generalized to the population (low external validity)

Question 10

archival research

Answer 10

• Gathering information through archival records (eg. Statistical reports, newspaper articles, speeches)
• Researchers use content analysis (the archival research equivalent of coding) to analyze documents and records (eg. Integrative complexity)
• 3 sources of data: statistical records (ex. GRE results), survey archives, written and mass communication records (ex. letters, diaries, speeches)
• may offer good generalizability, but may be hard to obtain, info collected by someone else may be inaccurate, and they can’t make casual claims

Question 11

observational methods

Answer 11

• naturalistic observation
• systematic observation
• case studies
• archival research

Question 12

ways of administering questionnaires

Answer 12

• in person
• mailed surveys
• internet-based
• other technologies/on-line measures (ie. cell phones, EAR)

Question 13

in-person questionnaires

Answer 13

• Most common form
• Easy and inexpensive to administer
• More easily control “test-taking” environment
• Requires easy access to target samples
• When asking participants to recall moods and memories, these memories may be biased (current mood may bias memories of previous week)

Question 14

mailed surveys

Answer 14

• Allows researchers to reach samples to which they do not have easy access
• More expensive than paper questionnaire
• No control over “test-taking” environment -> larger variance -> less power
• Very low return response rates
• When asking participants to recall moods and memories, these memories may be biased (current mood may bias memories of previous week)

Question 15

Internet-based

Answer 15

• Access to massive sample size, frees researchers from many physical constraints
• Very inexpensive
• No control over “test-taking” environment -> larger variance -> less power
• Difficult to verify authenticity and demographics
• When asking participants to recall moods and memories, these memories may be biased (current mood may bias memories of previous week)

Question 16

Questionnaires using other technologies

Answer 16

• Cell phones: texting or contacting people through an App and getting them to provide immediate report of current moods, activities, emotions, etc. (eliminates problem of recalling moods later -> bias)
• EAR = Electronically Activated Recorder: Small clip-on microphone and digital sound recorder -> at every pre-sent interval, EAR records 30 seconds of ambient sound -> assistants transcribe, then analyze, recordings

Question 17

2 broad categories of sampling techniques

Answer 17

• probability sampling

- non-probability sampling

Question 18

probability sampling

Answer 18

• Participants randomly drawn
• Likely representative of population
• High generalizability
• May cost more to do
• Types of probability sampling: simple random sampling, cluster sampling, stratified random sampling

Question 19

non-probability sampling

Answer 19

• Participants not drawn at random
• Easy and convenient
• Easy to introduce bias; low generalizability (not representative of population); very susceptible to self-selection
• types of non-probability sampling: convenience sampling, purposive sampling, quota sampling, snowball sampling

Question 20

simple random sampling

Answer 20

• type of probability sampling
• Everyone in a population has equal chance of being chosen (ex. Putting everyone’s names in a hat and pulling out names at random)
• Can be difficult to accomplish if your population is quite large (ie. Vancouver)

Question 21

cluster sampling

Answer 21

• type of probability sampling
• If “clusters” exist within a population, randomly sample clusters rather than individual members (ex. In Psyc 217, the clusters would be lab groups; at UBC, clusters would be majors/departments, etc.)

Question 22

stratified random sampling

Answer 22

• type of probability sampling
• Identify subgroups within population
• Identify proportion of people in population that belong to subgroups
• Randomly draw samples from subgroups to reflect proportion in population
• Ex. If demographics of Psyc 217 are 10% Caucasian men, 58% East Asian women, etc. Your sample would have to be comprised of 10% Caucasian men and 58% East Asian women

Question 23

convenience sampling

Answer 23

• type of non-probability sampling
• aka: haphazard sampling
• Recruit participants whenever and wherever you can (which may not give you the full scope of the population)
• Ex. Trying to study UBC students but only recruiting outside of International House

Question 24

purposive sampling

Answer 24

• type of non-probability sampling
• Targeting only people who fit a certain criterion or set of criteria, and recruiting them for a study
• Ex. Targeting people who want to stop smoking

Question 25

quota sampling

Answer 25

• type of non-probability sampling
• Similar to stratified sampling, but without the randomness
• Participants are selected using convenience sampling

Question 26

snowball sampling

Answer 26

• type of non-probability sampling

- Asking participants to recommend others to participate, helping researchers accumulate participants

Question 27

random assignment vs. random sampling

Answer 27

• Random assignment: every participant has an equal chance of being in any experimental condition in a study -> Allows for high internal validity
• Random sampling: every participant in the population has an equal chance of being selected to be in the sample -> allows for high external validity
• Is our random sampling ever truly random? Whole population -> people you can find -> people who say yes -> people who follow through

Question 28

interviews

Answer 28

• Response rates tend to be higher than questionnaires
• Rapport between interviewer and participant may motivate participant to answer more questions rather than leaving some blank on a questionnaire
• Interviewer can ask follow-up questions or clarify any confusing questions
• Problem: interviewer bias (could subtly influence answers by showing approval or disapproval, personal characteristics of interviewers may influence responses, interviewer may have expectations and “see what they are looking for” in respondent’s answers)
• types of interviews: face-to-face, telephone, focus group

Question 29

face-to-face interviews

Answer 29

• Interviewer may travel to respondent’s home or work to conduct interview
• Time-consuming and expensive -> used when sample size is small and there would be clear benefits to face-to-face interaction

Question 30

telephone interviews

Answer 30

• Used for large-scale surveys

- Less expensive and allow data to be collected relatively quickly

Question 31

focus group interviews

Answer 31

• Group of 6-10 people brought together for a few hours and usually given a monetary incentive to participate
• Open-ended questions asked to the whole group -> group interaction is an advantage of this method
• Each focus group is time-consuming and costly, but provides a lot of information

Question 32

population

Answer 32

set of people of interest to the researcher (ex. Eligible voters in Canada)

Question 33

sampling

Answer 33

obtaining information from respondents, which will be used to estimate the characteristics of the population as a whole

Question 34

external validity

Answer 34

• ensuring that the sample is representative of the population
• Random sampling helps achieve this

Question 35

sampling frame

Answer 35

• actual population of people from which a random sample will be drawn
• ex. If your population is residents of Vancouver, your sampling frame could be a list of home telephone numbers in Vancouver
• This can introduce bias – ex. Excludes people with unlisted numbers, without a home phone, etc.

Question 36

response rate

Answer 36

• survey of the percentage of people in the sample who actually completed the survey
• Ex. If you mail out 1000 questionnaires and 500 are filled out and returned, response rate = 50%
• The lower the response rate, the greater the likelihood that bias will distort the findings and limit generalizability -> low external validity